Predicting EHL film thickness parameters by machine learning approaches  

作　　者：Max MARIAN Jonas MURSAK Marcel BARTZ Francisco J.PROFITO Andreas ROSENKRANZ Sandro WARTZACK 

机构地区：[1]Department of Mechanical and Metallurgical Engineering,School of Engineering,Pontificia Universidad Católica de Chile,Santiago,6904411,Chile [2]Engineering Design,Friedrich-Alexander-University Erlangen-Nuremberg(FAU),Erlangen,91058,Germany [3]Department of Mechanical Engineering,Polytechnic School of the University of São Paulo,São Paulo,05508-030,Brazil [4]Department of Chemical Engineering,Biotechnology and Materials(DIQBM),FCFM,Universidad de Chile,Santiago,8370456,Chile

出　　处：《Friction》2023年第6期992-1013,共22页摩擦（英文版）

基　　金：support from Pontificia Universidad Católica de Chile.A.Rosenkranz gratefully acknowledges the financial support given by ANID(Chile)in the framework of the Fondecyt projects(Nos.11180121 and EQM190057);Additionally,A.Rosenkranz acknowledges the financial support given by the VID of the University of Chile within the project U-Moderniza(No.UM-04/19).

摘　　要：Non-dimensional similarity groups and analytically solvable proximity equations can be used to estimate integral fluid film parameters of elastohydrodynamically lubricated(EHL)contacts.In this contribution,we demonstrate that machine learning(ML)and artificial intelligence(AI)approaches(support vector machines,Gaussian process regressions,and artificial neural networks)can predict relevant film parameters more efficiently and with higher accuracy and flexibility compared to sophisticated EHL simulations and analytically solvable proximity equations,respectively.For this purpose,we use data from EHL simulations based upon the full-system finite element(FE)solution and a Latin hypercube sampling.We verify that the original input data are required to train ML approaches to achieve coefficients of determination above 0.99.It is revealed that the architecture of artificial neural networks(neurons per layer and number of hidden layers)and activation functions influence the prediction accuracy.The impact of the number of training data is exemplified,and recommendations for a minimum database size are given.We ultimately demonstrate that artificial neural networks can predict the locally-resolved film thickness values over the contact domain 25-times faster than FE-based EHL simulations(R^(2) values above 0.999).We assume that this will boost the use of ML approaches to predict EHL parameters and traction losses in multibody system dynamics simulations.

关 键 词：machine learning elastohydrodynamic lubrication film thickness support vector machine Gaussian process regression artificial neural network 

分 类 号：TG174.4[金属学及工艺—金属表面处理] TP183[金属学及工艺—金属学]